Company Profile
Nanjing Wotian Technology Co., Ltd. (referred to as "Wotian Technology"), founded in 2005, is a National High-Tech Enterprise and a National-Level Specialized, Refined, Differential, and Innovative (SRDI) "Little Giant" Enterprise. It is also the drafting unit for four national standards on pressure sensors. The company focuses on the R&D and manufacturing of pressure sensors, mastering the core technologies of diffused silicon and single-crystal silicon pressure sensors. In 2025, its annual production of pressure products will exceed 4.5 million units, with products exported to more than 80 countries and regions.
Editor's Note
When most people discuss semiconductors, their focus is on chips and advanced process nodes, rarely paying attention to pressure sensors that cost only a few dozen yuan each. Small as the device is, the consequences of its failure cannot be underestimated: abnormal braking perception in automobiles, uncontrolled cabin pressure in rockets, and failure of fire water pressure alarms. Behind these risks, it is this inconspicuous component that serves as the safety net. In this issue, we interview Zhao Jianli, Chairman of Wotian Technology. He has been deeply engaged in the entire pressure sensor chain and once served as the head of pressure detection for the national Project 921 aerospace program, witnessing the process of domestic sensors from aerospace breakthroughs to civilian mass production. If aerospace sensors can go to space, why is it harder to achieve large-scale automotive deployment? After 30 years of catching up, with R&D capabilities steadily advancing, why is industrial deployment still difficult? This issue will give you a clear picture of the reliability battle behind basic components.
*The following is a transcript of the podcast, confirmed and authorized by the guest.
Ubiquitous Pressure Sensors
Huanshi: First, let me ask a very basic question on behalf of our listeners. We often talk about automobiles, blood pressure monitors, gas meters, water pipes, oil pipes, as well as chemical, petroleum, exploration wells, and rockets. Pressure sensors are installed in all these places, but people usually overlook them. I would like to ask you to explain in the most accessible terms: this small device for measuring pressure sounds simple in principle, but what are the actual difficulties? And how can we ensure its stability and error-free operation?
Zhao Jianli: The core function of a pressure sensor is to sense the pipeline pressure of gases and liquids, and convert this physical pressure signal into a transmittable electrical signal. In the past, instruments could only display pressure values without signal output capabilities. With the advent of sensors, they can not only sense pressure but also transmit data outward. Its functions are mainly reflected in three aspects. In terms of application scenarios, first, it can support automation; second, it can save energy; and third, it can trigger abnormal alarms. Pipeline pressure itself has a safety threshold, and the sensor acts as a conversion carrier that transforms the pressure signal into a signal recognizable by the control system.
Huanshi: Where does the difficulty lie in making such a device?
Zhao Jianli: In many scenarios, uncontrolled pipeline pressure can lead to safety accidents such as explosions, so the primary requirement for sensors is reliability. For example, in the past, fire water pressure had to be manually checked from morning to night every day. Now, sensors can monitor it at any time. As long as the fire water runs out or the water pressure is insufficient, the sensor will quickly trigger an alarm, ensuring fire safety.
In addition, in livelihood scenarios, household water used to rely on water towers, which posed a problem: when residents used water simultaneously, there would be insufficient water pressure. For instance, in the 1990s, during the commercial breaks of TV dramas, a large number of residents used water at the same time, leading to water outages in toilets. But now such situations have disappeared, and it is precisely the sensors working behind the scenes. Modern high-rise buildings are equipped with a set of variable-frequency water pumps every three floors, monitoring the pipeline network pressure through pressure sensors: when the pressure is low, the pump power is increased; when the pressure is sufficient, the power is reduced. In this way, energy saving is achieved while ensuring stable water pressure, and residents no longer need to rely on water towers. Naturally, the scenario where centralized household water use leaves everyone without water is a thing of the past. This is the role of sensors.
The same applies to air conditioners—the "one kilowatt-hour per night" promoted by air conditioners actually relies on the regulation of sensors. By monitoring the status of the refrigerant (such as Freon, difluoromethane, etc.) through pressure sensing, the air conditioner does not need to start and stop repeatedly but can continuously and dynamically adjust. This is the role of cascade control, which can reduce the energy consumption of air conditioners by up to over 60%.
Cascade Control: It is a two-layer closed-loop control system. The outer loop (main loop) takes the target temperature as the control objective; the inner loop (secondary loop) quickly responds to various disturbances such as pressure and flow, completing the correction before the disturbances affect the room temperature. The pressure sensor is a key sensing component in the inner loop, providing real-time pressure data for the cascade algorithm.
Aerospace-Grade Sensors: Going to Space is Easier, but Automotive Mass Production is Harder
Huanshi: Mr. Zhao, how many years have you been researching pressure sensors? Your resume includes a rather special experience as the head of pressure detection for the national Project 921 aerospace program. Did this position bring you particularly great challenges?
Zhao Jianli: Strictly speaking, I started working on pressure sensors in 1991. From chips to packaging, and later to debugging, and now to applications, I have experienced every industrial link in the entire chain.
Project 921 belongs to the Shenzhou series of spacecraft projects. Starting from Shenzhou 1, we were responsible for the cabin pressure sensors. The role of the sensor is to maintain the cabin pressure after the launch of Shenzhou, ensuring the normal work of astronauts and equipment. For aerospace-grade sensors, reliability is extremely important. Because once in space, it is difficult to maintain or replace them, and there are still many unknowns in the space environment. When we first started our research, everyone was not very familiar with the space field. Under such circumstances, ground simulation was required, and only after passing the simulation tests could the products be put into use.
Huanshi: How high does the "reliability" you mentioned need to be? For such aerospace-grade micro devices, what is the magnitude of the difference compared to civilian products?
Zhao Jianli: Actually, sensors belong to high-end products and need to undergo reliability screening through multiple environmental tests such as vibration. Industrial, aerospace, and automotive sensors are all high-reliability categories, directly related to equipment and personal safety, requiring long-term stable operation without failure for several or even more than ten years.
In the automotive industry, the failure rate benchmark is 20 PPM, with leading companies targeting 10 PPM, meaning that out of one million products, the number of defective ones must be controlled within 10. When a car is running at high speed, sensors play a critical role. Once the pressure sensors in the braking, power, or battery systems fail, it will cause huge risks.
PPM: A manufacturing quality metric, representing the number of defective products per million. Example: 10 PPM means that for every 1 million products, a maximum of 10 defective products are allowed.
Achieving the above indicators requires three things: first, designing a reliable solution; second, strict control during the production process; and third, dedicated verification equipment. Claiming a ten-year lifespan for a product cannot rely solely on theory; it must be supported by data from laboratory accelerated life tests to ensure its long-term safety in use. Because when people buy a car, they don't use it today and replace it tomorrow.
Unlike aerospace products that can invest more resources in testing, achieving safety, reliability, consistency, and mass production for industrial products requires a lot of effort.
Huanshi: In the scenarios you just mentioned, such as automobiles, air conditioners, and aerospace, the differences in temperature, corrosion, and vibration environments faced by pressure sensors are very significant. In these scenarios, it is not easy to achieve product reliability. In your opinion, what is the biggest challenge in the multi-scenario development process?
Zhao Jianli: Depending on different scenarios, products need to be custom-designed. The development logic is divided into three steps: first, understanding the application scenario; second, controlling the production process; and finally, verification through testing.
Actually, "reliability" itself is the most difficult point, and every step is not easy. Because the entire product process may involve hundreds of processes, the defect rate of each process must be compressed to the PPM level. Loss of control in any link will bring the risk of complete system failure.
Speaking of this, many people have a misconception that aerospace and special products are the hardest to make. Actually, from my personal perspective, making pressure sensors for special and aerospace applications is far less difficult than making them for automobiles. Because special and aerospace products have small batches, single-product screening can be done, and technical personnel can follow up on problems throughout the process. But when producing 1 million or 10 million automotive pressure sensors, it is very difficult to maintain an extremely low failure rate while controlling costs. This is the most difficult point; the core of large-scale industrialized manufacturing is to stably produce products with reliable quality.
Huanshi: Indeed, the production batch of most special and aerospace products is relatively small, and the gross profit per unit is high enough to support the process cost of piece-by-piece screening. But the automotive industry is different. With a shipment scale of tens of millions, the single-product screening model cannot be followed, which poses a higher challenge to the yield rate.
From Chips to Packaging: The Trust Dilemma of Domestic Sensors
Huanshi: Twenty years ago, many people in the industry chose to buy overseas pressure sensors for packaging, but you did the entire chain of chips, packaging, and testing. What was the opportunity that led you to choose this path at that time? And is the domestic sensor sector still being bottlenecked by overseas players now?
Zhao Jianli: The opportunity at that time was the difficulty of market deployment. At the beginning, we wanted to solve the chip problem. But after making the samples, the market did not recognize them, so we could not sell them. We judged that just making chips could not solve customers' pain points, so we continued to extend downstream and took up the packaging process as well.
Sensors are a marginal interdisciplinary field, integrating technologies in IC (Integrated Circuit) processes, materials, sintering, and welding. During the R&D process, we encountered many basic material difficulties and process problems, which could only be solved one by one. In this process, we went from making semi-finished products to finished products, and finally developed a series of products.
Huanshi: Did the logic of doing chips and packaging together solve the downstream customer recognition problem you mentioned at that time?
Zhao Jianli: The product deployment method at that time was generally for customers to try them out first. When the customers' pain points were solved, they naturally accepted them gradually. In this process, the gap between domestic chips and foreign products became smaller and smaller, and the introduction was finally completed.
Relatively speaking, making chips is not as difficult as packaging. Because chip production relies on processes and equipment, as long as the supporting equipment is in place, the difficulty of advancement is relatively controllable. When we first started making chips, even if the equipment was not complete, we could develop our own equipment to complete tape-out. However, packaging involves multiple interdisciplinary fields; while chip design essentially focuses on a single disciplinary field.
Huanshi: This is also the reason why the IDM (Integrated Device Manufacturing) model is relatively common in the sensor industry.
IDM Model: A semiconductor vertical integrated manufacturing model where enterprises independently complete the entire chain of chip design, wafer fabrication, packaging, testing, and finished product application.
Zhao Jianli: Yes, overseas companies started making sensor products earlier, and the IDM model has been verified more fully. If domestic enterprises directly draw on this experience, it is relatively simple to get started.
However, high-end sensors are generally subject to export restrictions. The global high-end sensor market is highly concentrated. Relying on the early application accumulation in the aerospace and special industries, American companies occupy most of the high-end market. Sensors belong to underlying basic components. If there are problems in the supply chain, it will directly affect the operation of factories and critical infrastructure.
At present, sensors for ordinary domestic scenarios have achieved autonomy, but the high-end market is still dominated by overseas products. Domestic enterprises are still in the catching-up stage.
Thirty Years of Catching Up: Application Verification is the Key Hurdle
Huanshi: The domestic industry has been catching up for 30 years. What is the current level compared to overseas?
Zhao Jianli: Early domestic development was relatively backward. In the last 10 years, domestic enterprises have continuously increased R&D investment in high-end products. At the level of R&D and design, the difference is no longer very large. The current core shortcoming is not in design capabilities, but in the accumulation of application scenario verification.
Application deployment itself is a process of continuous trial and error, requiring a lot of time and cost investment, which is also the difficulty we need to overcome at present. On the one hand, enterprises need to build brands; on the other hand, the core goal of enterprises is to create and complete product verification. Therefore, the industry is continuously investing, focusing on two major sectors: R&D and verification. We have built our own environmental laboratories to carry out various environmental reliability tests, exposing problems in continuous operation tests, and continuously reducing product defects. This is exactly what we are all doing at present.
Speaking of the overseas situation, it is actually the same as domestic aircraft. Whether it is special aircraft or civil aircraft, once overseas entities implement supply restrictions and stop supplying, it will instead force the local industry to accelerate its growth. For us, this is also an opportunity. Relying on the complete domestic manufacturing base, high-end chips and sensors like ours strictly speaking started R&D twenty years ago, but at that time, the application scenarios for deployment were limited. In recent years, the application market scale has grown rapidly, and a large number of deployment opportunities have emerged, bringing rare development opportunities.
Huanshi: One last question. Looking back at history, what message would you like to give to yourself when you first entered the industry, or to the young people who are just entering the industry now?
Zhao Jianli: Every era is different, and living conditions are also different. I think it is necessary for young people to pursue innovation. But what do I think is more important? It is to do one thing and do it well, and to persevere. We have been in this industry for more than thirty years. If we persist, it is easier to achieve results.
Huanshi: You mean that everyone still needs to endure the "cold bench".
Zhao Jianli: It's not really about enduring a "cold bench," because most people nowadays are not living just for tomorrow's food and clothing.
The current era is a good era, and young people have time to do what they like. In addition, they have the opportunity to access global cutting-edge technologies and have more room for innovation. Young people should be happy, and as long as they persist, they will continuously receive positive feedback.
Huanshi: Alright, thank you, Mr. Zhao, for being our guest on "Chip Unveiled" today. We also wish Wotian Technology's pressure sensors success in expanding into various scenarios and directions!
Zhao Jianli: Thank you.
Interview | Huanshi Editor | Xiaoqie Review | Huanshi